Living with spreadsheets Dean Buckner Financial Services Authority JULY 2011

AGENDA Recap on last year’s talk –Why we won’t get rid of spreadsheets But how can we live with them?

Why we won’t get rid of spreadsheets The tower of Babel Early views on machine translation (and why they failed) The computer Babel

The tower of Babel “And the whole earth was of one language, and of one speech. “And they said, Go to, let us build us a city and a tower, whose top may reach unto heaven; and let us make us a name, lest we be scattered abroad upon the face of the whole earth. “And the Lord said, Behold, the people is one, and they have all one language; and this they begin to do; and now nothing will be restrained from them, which they have imagined to do. “Go to, let us go down, and there confound their language, that they may not understand one another's speech. “Therefore is the name of it called Babel; because the Lord did there confound the language of all the earth.”

Machine translation Proposals for mechanical translators of languages pre-date the invention of the digital computer. The first recognisable application was a dictionary look-up system developed at Birkbeck College, London in 1948.

Code breaking Warren Weaver had been involved in code-breaking during the Second World War. A simple idea: given that humans of all nations are much the same (in spite of speaking a variety of languages), a document in one language could be viewed as having been written in code. Once this code was broken, it would be possible to output the document in another language. From this point of view, Chinese was English in code. “… one naturally wonders if the problem of translation could conceivably be treated as a problem in cryptography. When I look at an article in Russian, I say: "This is really written in English, but it has been coded in some strange symbols. I will now proceed to decode."

It failed US funding of Machine Translation research cost the U.S. public $20 million by the mid 1960s. The Automatic Language Processing Advisory Committee (ALPAC) produced a report on the results of the funding and concluded that "there had been no machine translation of general scientific text, and none is in immediate prospect".

It failed again? There was renewed interest in the 1980s with the emergence of the ‘artificial intelligence’ idea. (At least if Google translator is anything to go by) –Seinen Lebensabend verbrachte in bad kleinen, in der Nähe seiner Geburtsstadt Wismar. –His life was spent in small bathroom, near his hometown of Wismar.

Why it is difficult The teacher sent the boy to the headmaster because –he wanted to see him –he had been throwing stones –he was fed up with his bad behaviour

The computerised Babel In the beginning was the mainframe –Keep the ‘meaning’ of every symbol in just one place, and have everything else inside the system point to it directly (a ‘pointer’ is simply a mechanical means of moving from one address to another’) –Force users either to check their translation by means of a ‘compiler’ (this is for users called ‘programmers’) –or have them enter information by means of a menu system that forces acceptable choice (for common or garden users). –This worked reasonably well until the 1990s

The tower crumbles The 1980s and 1990s saw increasing specialisation of systems –General ledger systems –Payment systems –Loan systems –Claims systems etc They couldn’t talk to each other 

The modern Babel A modern bank or insurance company contains dozens, perhaps hundreds of disparate systems. There is no ‘compiler’ to allow communication between them Spreadsheets are the solution to this communication problem

Deceptively difficult problems Deceptively difficult problem: a problem whose solution seems easy –particularly by the application of ‘technology’ But isn’t As we saw, communication between systems is incredibly difficult –not like ‘code-breaking’ at all But it seems easy –I say: "This is really written in English, but it has been coded in some strange symbols. I will now proceed to decode."

Apparently easy solutions (1) The Internet –The Internet became embedded in popular consciousness in the 1990s and 2000s –The problem of sending data from one place to another seemed to be solved –But it didn’t solve the communication problem –The Chinese send a letter to English speakers, who receive it OK. But no one understands it.

Apparently easy solutions (2) Data warehouses –An apparently simple solution –Send all the data from disparate source systems into one place (the ‘warehouse’) –Then you have it all in one place But the problem remains – you have all the different languages in one room –And no one understands each other –Even worse, when the translation was done on spreadsheets, at least the users understood what was going on –Now nobody does

Large spreadsheet systems Spreadsheet systems are becoming huge –We saw a 600 spreadsheet system last year. That seemed big –Then we saw a 1,000 sheet system. That was even bigger. –Then we found a 9,000 sheet system. That was awesome. What do we do?

Dangers of large systems Large spreadsheet systems are like mainframes –But they don’t have a central compiler –The embedded risks are huge

Examples Hard-coded references passing unchecked through many spreadsheets –Date, source, and type of data is completely opaque –Nature of transformations completely unclear. –Location of transformations unknown

Examples Senior management sees only immediate source sheets –Under a dozen seems manageable –But they don’t see the hundreds or thousands of sheets that are feeding the dozen. –Tip of iceberg

Solving the problem [this page deliberately left blank]

Questions & Comments